Skip to content

[GitHub Trending] huangruiteng/loopx

8.1 relevance
Score Breakdown
technical depth
9
novelty
9
actionability
7
community
6
strategic
7
personal
9

Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.

Lightweight loop engineering state kernel for long-running AI agents, highly novel and directly relevant.

AI/ML github.com
Lightweight loop engineering state kernel for long-running AI agent teams. Agent-loop agnostic across Codex, Claude Code, and other coding agents, with durable goals, quota-aware auto-wake, executa...
Summary

LoopX is an open-source local control plane for long-running AI agent work, providing a lightweight state kernel that keeps objectives, gates, todos, evidence, quota, and handoffs stable across multiple agent runtimes like Codex, Claude Code, and Cursor. It implements an agent-native Kanban model where cards carry identity, authority, and continuation, with validated operators for claim, gate, monitor, and writeback, enabling multi-day engineering, research, and peer-agent team workflows without replacing the agent runtime itself. The project demonstrates real-world use through 200+ hour trajectories like OpenViking issue-fix and Auto ML experiments, emphasizing human judgment for dangerous permissions and production writes.

Author

huangruiteng